The alkaline province of southern West Greenlandincludes swarms of dykes described as kimberlites andlamproites (Larsen 1991), and these rock types arewidely distributed in the Sisimiut–Sarfartoq–Kanger-lussuaq region (Figs 1, 2). Kimberlites and lamproitesare potential carriers of diamond, and since the descrip-tion of the Sarfartoq carbonatite complex and the kim-berlitic dykes related to this complex (Larsen 1980;Secher & Larsen 1980), the Sisimiut–Sarfartoq–Kanger-lussuaq region has seen several campaigns of com-mercial diamond exploration. The latest and mostpersistent stage of exploration began in the mid-1990sand has continued to date, with varying intensity.Numerous reports of diamond-favourable indicator min-

erals from till sampling, finds of kimberlitic dykes, andrecovery of actual diamonds from kimberlitic rockshave emerged since 1995 (Olsen et al. 1999). A drillingprogramme in late 2001 confirmed the unusually greatlength and width of a magnetic kimberlitic dyke(Ferguson 2001).

The alkaline ultramafic dykes within the Sisimiut–Kangerlussuaq and Sarfartoq regions intrude the bor-der zone between the Archaean craton and the Palaeo-proterozoic Nagssugtoqidian orogen (van Gool et al.2002, this volume). This border is defined as the south-ern boundary of Palaeoproterozoic reworking of theArchaean basement gneisses. The reworking has affectedthe Palaeoproterozoic Kangâmiut dolerite dykes, which

57

Kimberlites and other ultramafic alkaline rocks in theSisimiut–Kangerlussuaq region, southern West Greenland

Sven Monrad Jensen, Henriette Hansen, Karsten Secher,Agnete Steenfelt, Frands Schjøth andThorkild M. Rasmussen

Archaean

Palaeoproterozoic

Undifferentiated

Arfersiorfik quartz dioriteSisimiut charnockite

Syntectonic granite suite

Archaean gneiss (s.l.)

Supracrustal rocksMafic dykes

Early Archaean gneiss

Kangâmiut dyke swarm

ThrustHigh strain zone

Ikertooq

Sisimiut

NordreIsortoq

SukkertoppenIskappe

InlandIce

Itilleq

Kangaamiut

Maniitsoq

NordreStrømfjord

Ikertooqsteep belt

52°54°

50 km

Søndr

e Strø

mfjo

rd

67° 67°

66°

Greenland

Fig. 1. Geological map of the region withthe project area (Fig. 2) outlined in a redframe. Modified from Mengel et al. (1998).

Geology of Greenland Survey Bulletin 191, 57–66 (2002) © GEUS, 2002

GSB191-Indhold 13/12/02 11:30 Side 57

were intruded into the Archaean gneisses prior to defor-mation and are now highly strained and boudinagednorth of the boundary.

The intrusion of alkaline dykes appears to have takenplace during two events. Lamproite dykes in the Sisimiutregion are around 1.2 Ga old, whereas the Sarfartoq com-plex (Fig. 2) and a swarm of associated kimberliticdykes have ages of around 0.6 Ga (Larsen & Rex 1992).A precise spatial relationship between the two intrusiveevents has not been established.

A project concerned with general scientific aspectsof kimberlitic and related rocks in the Sisimiut–Kangerlussuaq region was established in 2000. The aimis to gather knowledge from companies’ explorationactivities and to incorporate Survey expertise such aspetrology, structural geology, isotope geology, geo-chemistry, geophysics and the geographic informationsystem (GIS). Field work was carried out in 2000 and2001, with the Kangerlussuaq International ScienceSupport (KISS) centre at Kangerlussuaq airport as logis-tical base. A combination of daily reconnaissance cov-ering large areas using a helicopter, and fly camps of2–3 days duration was applied.

The broad term ‘kimberlitic’ is used here in acknow-ledgement of the fact that the classification of the ultra-mafic dyke rocks of West Greenland is not resolved

with unanimity. Many of the dyke rocks resemble kim-berlites and have previously been described as such(Larsen 1980, 1991; Scott 1981; Larsen & Rex 1992).Mitchell et al. (1999), however, take the view that thedyke rocks are not ‘archetypal’ kimberlites, but ultra-mafic lamprophyres that should be termed melnoitesor aillikites. In the light of the actual occurrence of dia-monds in the West Greenland dyke rocks, Mitchell etal. (1999) consider them to represent one of the fewexamples of diamond-bearing ultramafic lamprophyres.One of the aims of the present project is to contributeto a generally accepted classification of the WestGreenland kimberlitic and related rocks.

Distribution of kimberlitic and related rocksLarsen’s (1991) compilation of Greenland’s kimberlite,lamproite and ultramafic lamprophyre occurrencesproved to be a valuable initial guide to companies whenthe ongoing diamond exploration commenced in WestGreenland in the mid-1990s (Olsen et al. 1999). The com-pilation includes data collected by earlier workers (e.g.Scott 1981), who typically operated by boat and thuswere often restricted to working near the coast. As a

58

0 40 km20

67°

66°30´

53° 52° 51° 50°

•

•Kangerlussuaq

Inland Ice

•Sisimiut

Sarfartoq

Sønd

reStr

ømfjo

rd

A

B

C

ED

F

G

H

I

J

Itilleq

Fig. 2. Distribution of kimberlitic and lamprophyric occurrences within the project area. A–J are localities mentioned in the text. Redtriangles are kimberlitic and lamprophyric dykes known prior to the present study: lamproites predominate in the region betweenSisimiut and Itilleq, kimberlites in the Sarfartoq region (Larsen & Rex 1992). Green triangles are observations from the present study.Heavy colouring represents in situ finds and light colouring represents boulders (float).

GSB191-Indhold 13/12/02 11:30 Side 58

consequence, the known occurrences had an unevendistribution, and some areas appeared to be withoutoccurrences.

Many of the occurrences of in situ dykes and boul-ders (float) in the Sisimiut–Sarfartoq–Kangerlussuaqregion known at present are indicated in Fig. 2. Theapproximately 600 occurrences include those fromLarsen’s (1991) compilation and new localities addedby exploration companies and recent field work by theGeological Survey of Denmark and Greenland (GEUS).It should be noted that the map does not include alarge number of company localities that are at presentconfidential because they lie within active exclusiveexploration licence areas. The map illustrates the con-centration of occurrences in the vicinity of the Sarfartoqcarbonatite complex, which Larsen (1980) interpretedas a swarm of kimberlitic cone-sheets centred on thecarbonatite.

Bedrock exposure is generally poor in the studyregion, and dykes and sills of kimberlite, lamproite andcarbonatite can be difficult to trace because they arethin, easily eroded and often covered by overburdenand vegetation. Therefore, both stream geochemicaldata and aeromagnetic data are investigated with theaim of identifying geochemical and geophysical signa-tures reflecting occurrences of kimberlitic and relateddykes.

Field activitiesField work in 2000 consisted of one week’s reconnais-sance and visits to new dyke occurrences discovered

by exploration companies. The 2001 field work wasfocused on the spatial distribution of kimberlitic dykerocks in areas with little or no previous information, anddetailed studies on mantle xenoliths from the kimber-litic dykes. Studies of mantle xenolith-bearing rocks aredescribed in a separate section below. The field workalso included testing the ease of locating occurrencesdescribed in companies’ assessment reports using var-ious Global Positioning System (GPS) equipment andGIS methods.

Locality F (Fig. 2) provides an example of the den-sity of new observations made in an area with no pre-viously reported kimberlitic or lamproitic occurrences.Within the framed area 26 kimberlitic boulders and onenew dyke were registered (Fig. 3). Three of the boul-ders measure about 1 x 1 x 1 m in size, indicating thatthe dyke or dykes from which they originated must beof substantial dimensions. Several large boulders werebroken up into many smaller fragments, resulting in amuch higher count of individual pieces of kimberliticrock. The high number of kimberlitic boulders in thissmall area suggests they may have been derived froma nearby dyke outcrop. Similar observations were madeat localities G and H (Fig. 2).

Stream sediment geochemical anomalies reflecting alkaline rocksGeochemical reconnaissance-scale stream sediment sur-veys have been carried out over large parts of WestGreenland, including the project region consideredhere. The chemical analyses have been compiled and

59

Fig. 3. View of the area of locality F. Asmall outcrop of a new kimberlitic dyke(arrow) is located in the lineamentmarked by dotted line. Length of dottedline is approximately 200 m.

GSB191-Indhold 13/12/02 11:30 Side 59

calibrated (Steenfelt 1999, 2001a), and presented in ageochemical atlas of West and South Greenland (Steenfelt2001b). The atlas displays element distribution patternsfor 43 major and trace elements based on analysis ofthe < 0.1 mm grain size fraction of stream sediment sam-ples collected systematically with a density of aboutone sample per 30 km2. The atlas also includes mapsof high-pressure mineral phases of assumed mantlederivation, so-called kimberlite indicator minerals, iden-tified in the 0.1 to 1 mm fraction of stream sedimentsamples collected between latitudes 61°N and 67°N.

The Sarfartoq carbonatite complex comprises rocksenriched in P, Ba, Sr and rare-earth elements (REE),and hosts a niobium-rich pyrochlore mineralisation(Secher & Larsen 1980). In the geochemical maps, the

Sarfartoq complex is easily recognised as a local but verypronounced anomaly with high concentrations of P,Ba, Sr, Nb and REE, as well as Mo, Ta and Th.

Kimberlites and lamproites have olivine- and car-bonate-rich groundmass (high Mg and Ca) and highconcentrations of the trace elements Ba, Cr, Ni, Nb, Sr,P, Ti and light REE. In view of the particular minera-logy and chemistry of the alkaline rocks, their presencewithin the gneiss-dominated Archaean basement shouldbe readily detectable in the stream sediment geochemicalsurvey. However, because of the small volume of alka-line rocks, their geochemical imprint on the stream sedi-ment composition may be obscured by the presenceof other igneous rocks that have intruded the tonaliticgneisses in the same region. Gabbroic inclusions withinthe gneisses, and the Palaeoproterozoic Kangâmiutdolerite dyke swarm, release so much Ni, Cr and Ti tothe streams that the contribution of these metals fromalkaline ultramafic rocks is easily disguised. Furthermore,Palaeoproterozoic intrusions within the Nagssugtoqidiandomain comprise considerable volumes of charnock-ite and syenite, enriched in Ba, Sr, P and REE (Steenfelt1997), which give rise to anomalies that are similar tothose expected to be caused by kimberlites and lam-proites.

However, one element, niobium, has proved con-vincing as a ‘pathfinder’ for kimberlites and lamproites.Table 1 illustrates the high Nb concentrations of kim-berlites and lamproites relative to Archaean gneisses;even the Palaeoproterozoic syenites have relatively low

60

Table 1. Range of Nb concentrationsof various rock types in the

Sisimiut–Sarfartoq–Kangerlussuaq region

Rock type Age Nb ppm

Gneiss Archaean 0.5–11Charnockite 1.9 Ga 1–25Syenite 1.9 Ga 5–25Lamproite 1.2 Ga 95Kimberlite 0.6 Ga 163–311Beforsite 0.6 Ga 151Carbonatite 0.6 Ga 2.3–22

From Larsen & Rex (1992) and GEUS database (unpublished).

Nbppm 0–1112–1516–24

24 >

53° 52° 51° 50°

67°

66°30´

0 40 km

★

▲▲

▲

▲

▲

▲▲

Sønd

reStr

ømfjor

d

ItilleqS

Fig. 4. Distribution of Nb in the < 0.1 mmfraction of stream sediments. S: Sarfartoq;black triangles: high-Mg, high-Crchromite; black diamonds: high-Cr,low-Ca pyrope garnet.

GSB191-Indhold 13/12/02 11:30 Side 60

contents of Nb. The regional stream sediment anomalypattern for Nb in the study region (Fig. 4) does, in fact,coincide with the province of known kimberlites andlamproites.

It was recognised at an early stage that the Sarfartoqcomplex and surroundings were characterised by aregional stream sediment Nb anomaly (Steenfelt 1991).The anomaly was then attributed partly to Nb miner-alisation, and partly to the known ultramafic alkalinedykes. As more stream sediment data became available(Steenfelt et al. 1993) and more kimberlite occurrenceswere registered, the apparent spatial relationshipbetween the high Nb in stream sediments and occur-rences of ultramafic alkaline dykes was strengthened.

Field visits to high-Nb stream sites were carried outin 2000 and 2001 to test the validity of high Nb in streamsediments as an indicator of kimberlite and lamproite.At one site with a very high Nb value a lamproite dykewas seen in outcrop in a stream gorge, and at other sitesboulders of kimberlite were observed along the streamcourse. Outcrops or boulders of kimberlite or lamproitewere not found in all the high-Nb sites visited. However,it is worthy of note that thin dykes and float of befors-ite have been recorded some distance from the Sarfartoqcarbonatite complex; beforsite is a magnesiocarbon-atite rock with high concentrations of Nb and LREE,which probably contributes to the Nb-enriched province.

Stream sediment samples with high Nb and high-pressure chromite have been collected in areas eastand south-east of known kimberlite occurrences sug-gesting a wider extent of the kimberlite field than recog-nised hitherto. A small cluster of high Nb values in thenorth-eastern part of Fig. 4 is at present unexplained.Kimberlites or lamproites have not so far been recordedhere. However, the generally poor exposure in thestream surroundings would hinder ready recognition ofalkaline dykes or other rocks with high Nb such asgranitic pegmatites.

High concentrations of P, REE, Ba and Sr are recordedin stream sediment samples collected in a 15 x 15 kmarea near the coast south of Itilleq (locality I, Fig. 2).The geochemical signature of these samples is similarto that obtained in streams draining Palaeoproterozoicsyenitic intrusives in the Nagssugtoqidian domain(Steenfelt 1998). It is noteworthy that the Nb concen-trations are as low as in the Nagssugtoqidian syenites,and hence it was not expected to find kimberlites, butsome other kind of alkaline rock.

Two field days were spent in a valley upstream fromthe most pronounced stream sediment anomaly. Thearea is dominated by a gneiss complex intruded by a

10 to 15 m wide Kangâmiut dyke and numerous thin,less than 0.5 m wide, dykes and sills of grey lamprophyreand a brown dolerite with star-shaped aggregates of phe-nocrystic plagioclase. Blocks of a coarse-grained alka-line ultramafic rock were frequently seen in the morainesof the valley glacier and were particularly abundant inthe screes covering the southern valley slopes. Thealkaline rock is dominated by black, shiny amphiboleand has variable to large amounts of brown phlogopiteand light green apatite. The finds explain the geo-chemical anomaly and enlarge the province of knownalkaline rocks in this part of Greenland.

Mantle xenoliths in kimberlitic rocksMantle xenoliths are widely used in studies of the man-tle lithosphere (Menzies & Hawkesworth 1987; Nixon1987, 1995), and are here used to shed light on howthe plate-tectonic history affected the deeper parts ofthe lithosphere across the Archaean–Proterozoic tran-sition in the Sisimiut–Sarfartoq–Kangerlussuaq region.An extensive collection of xenoliths from several kim-berlitic occurrences in the unreworked Archaean ter-rane south of Søndre Strømfjord was made in 2000,and two kimberlitic dykes containing abundant man-tle xenoliths (localities A and B, Fig. 2) were visited andsampled in 2001. At least 90% of the xenoliths encoun-tered have peridotitic or pyroxenitic compositions, andrange in size from less than 1 cm to 15 cm across. Somexenoliths contain visible purple pyrope garnet (Fig. 5)and green Cr diopside, minerals that – given the right

61

Fig. 5. Kimberlitic dyke with many different types of mantle xeno-liths (locality A, Fig. 2). Note purple garnets indicated witharrows; hammer and pencil for scale.

GSB191-Indhold 13/12/02 11:30 Side 61

chemical composition – may indicate P–T conditionswithin the diamond stability field.

In the Palaeoproterozoic reworked Archaean terranenorth of Søndre Strømfjord another kimberlitic dykereported by exploration companies was visited (localityC, Fig. 2). The dyke consists of a 5–10 m wide, c. 500 mlong train of kimberlitic boulders and outcrops in bandedgrey gneiss. Peridotitic and pyroxenitic xenoliths areabundant, ranging in size from less than 1 cm to around10 cm. Dunitic (ol), harzburgitic (ol + opx) and lher-zolitic (ol + opx + cpx ± gt) xenoliths were identifiedin the field. The Survey campaign located two furtherdykes north of Søndre Strømfjord (localities D and E,Fig. 2). The locality D dyke is vertical, 2–5 m wide andcan be traced for c. 1.5 km. Xenoliths are much lessabundant here than in the dykes south of Søndre Strøm-fjord, and their sizes range from less than 1 cm to about5 cm. Macroscopically, the peridotitic xenolith typesare dunite (ol) and lherzolite (ol + cpx + opx). The local-ity E dyke is a gently dipping (16–20°S), 30–50 cm thicksheet, and contains small xenoliths that are mostly lessthan 2 cm in diameter.

Preliminary resultsSeven xenoliths from the locality A dyke have beenanalysed for major element mineral chemistry by elec-tron microprobe at the Geological Institute, Universityof Copenhagen, and garnets in three of them have beensubjected to reconnaissance laser ablation analyses atthe Survey. Garnets in the same three xenoliths weresubjected to high-precision analysis of Ni by electronmicroprobe, and the xenolith rock types (Table 2) havebeen determined from their mineral assemblages andestimated mineral proportions (e.g. LeBas & Streckeisen1991).

Generally, the major element compositions of min-erals within individual xenoliths are homogeneous

(Table 2). The average forsterite (Fo) content of theolivines ranges from Fo87 to Fo92; garnet wehrlites con-tain the olivines with the lowest Fo, garnet harzburgitesand garnet lherzolites the olivines with the highest Fo.Average Ni contents are within the range of 2900 ± 360ppm reported by Ryan et al. (1996) for garnet peri-dotites. Orthopyroxenes contain less than 1.6 wt% CaOand have relatively low Al2O3, the lowest values beingfrom ilmenite-free garnet lherzolites (0.49–0.53 wt%).Clinopyroxene Mg/(Mg + Fe2+) ratios correlate posi-tively with Cr2O3. Garnet Mg/(Mg + Fe2+) ratios tend tocorrelate positively with Fo contents of olivine. CaO andCr2O3 contents of garnets show limited variation withinindividual xenoliths but large overall variations (Fig. 6),and the garnets from two harzburgite xenoliths clearly

62

Table 2. Calculated temperatures and pressures for xenolith minerals from the locality A dyke

T, P and TNi calculated according to: * Nimis & Taylor (2000), † Ryan et al. (1996). ol: olivine, opx: orthopyroxene, cpx: clinopyroxene, gt: garnet, ilm: ilmenite, hzb: harzburgite, lhz: lherzolite, weh: wehrlite.The GGU numbers refer to material from Greenland in the files of the Geological Survey of Denmark and Greenland.

Mineral GGU sample 464002.2 464003 464006 464008.10 464008.11analysed Rock type gt hzb gt lhz gt lhz gt lhz gt weh

Minerals ol-opx-gt ol-opx-cpx-gt ol-opx-cpx-gt ol-opx-cpx-gt-ilm ol-cpx-gt

cpx T (°C) * 1039 ± 30 1194 ± 30 1124 ± 30 1030 ± 30cpx P (kbar) * 48.9 ± 2.3 57.8 ± 2.3 68.7 ± 2.3 52.5 ± 2.3gt TNi (°C) † 1143 ± 42 1101 ± 18 1221 ± 30

0

2

4

6

8

10

12

0 2 4 6 8

CaO (wt%)

STD

464002

464003

464006

464008.8

464008.10

464008.11

464007.1

G9G10

Cr 2

O3

(wt%

)

Fig. 6. Garnet compositions in xenoliths from the locality A dyke,analysed by electron microprobe. The G9–G10 boundary line isfrom Gurney (1984) and Gurney & Zweistra (1995), and STD isa reference garnet analysed along with microprobe standards. GGU464xxx numbers refer to sample material in the files of theGeological Survey of Denmark and Greenland.

GSB191-Indhold 13/12/02 11:30 Side 62

plot within the G10 field of Gurney (1984) and Gurney& Zweistra (1995). G10-class pyrope garnets are con-sidered to be strongly indicative of conditions favourablefor diamond stability.

Equilibration temperatures (T) and pressures (P) forthe clinopyroxene-bearing xenoliths were calculatedusing the Nimis & Taylor (2000) single clinopyroxenethermobarometry equations (Table 2), and these tem-perature estimates are comparable to TNi values calcu-lated using the high-precision Ni analyses and theempirical Ni-in-garnet thermometer of Ryan et al. (1996).

Variations of Zr vs. Y in garnets of three xenoliths fromthe locality A dyke are shown in Fig. 7, and averagedchondrite-normalised REE patterns for garnet and clinopy-roxene in each xenolith in Fig. 8. The very low Zr andY contents in garnets of a garnet harzburgite xenolith(sample GGU 464002.2) are consistent with their low-Ca character (Fig. 6) and indicate that this xenolith rep-resents original lherzolite mantle depleted by partialmelting (Griffin et al. 1992). The garnet in sample GGU464002.2 is more depleted in the middle and heavy REEthan the garnets in lherzolites from samples GGU 464003and 464006. The sinuous pattern resembles that of low-Ca garnet in a harzburgite xenolith from the Sarfartoqarea (Garrit 2000). The low REE contents support theidea that the garnet is hosted in a rock depleted by par-tial melting, whereas the sinuous shape may indicate post-depletion metasomatic enrichment processes (Hoal et al.1984) or disequilibrium garnet growth (Shimizu & Sobolev1995). Garnet lherzolites from samples GGU 464003 and

464006 have much higher Zr and Y concentrations inthe garnets, and their chondrite-normalised REE pat-terns are more like those of typical LREE-depleted andMREE- to HREE-enriched garnets of primitive mantle(e.g. Haggerty 1995). Garnets from samples GGU 464003and 464006 plot within fields indicating phlogopite andmelt metasomatism of previously depleted mantle, respec-tively (Fig. 7). Phlogopite occurs in the xenoliths as partof kelyphitic rims on garnets, and in sample GGU 464003also as a minor phase away from garnets. Ti contentsof garnets, which follow Zr in some metasomaticprocesses (Griffin et al. 1999), show no significant zon-ing. This may indicate that if metasomatic processesaffected garnet compositions, there was sufficient timefor equilibration with the metasomatising agents.

Geophysical properties of kimberlitesAirborne geophysical surveys play an important role indiamond exploration, as kimberlites are often hiddenunder surficial deposits but also often have magneticproperties that make them distinguishable from thecountry rocks (e.g. Keating 1995; Macnae 1995). Since1996 several helicopter-borne surveys have been com-missioned by exploration companies to cover theirlicence areas in the Maniitsoq and Kangerlussuaq regions(Olsen et al. 1999). Data from the latest helicopter-borne survey in the Kangerlussuaq region in 2000 wereused to outline targets for the drilling that led to the

63

0

10

20

30

40

50

60

0 100 200 300 400

Zr (ppm)

Y (

ppm

)

Gt SEP

464002.2

464006

464003

Lh-1

Lh-2

Hz

Und

eple

ted

garn

ets

(hig

h ge

othe

rm)

Depleted garnets

Melt

met

asom

atism

Phlogopitem

etas

omat

ism

Fig. 7. Laser ablation ICP-MS analyses ofZr and Y in garnets from xenoliths in thelocality A dyke (this study, data fromGEUS), garnet lherzolite and garnetharzburgite xenoliths from the Sarfartoqarea (Garrit 2000, data from MemorialUniversity, Canada) and in garnetseparates from the Sarfartoq area(brown dots: Garrit 2000, data fromMacquarie University, Australia). Thecoloured fields are from Griffin & Ryan(1995) and Griffin et al. (1999), andarrows indicate core-to-rim zonationdirection in metasomatised garnets ofGriffin et al. (1999).

GSB191-Indhold 13/12/02 11:30 Side 63

confirmation of a 5 km long and 20 m wide kimberliticdyke in late 2001 (locality J, Fig. 2; Fig. 9; Ferguson 2001).

The petrophysical properties of 22 representative kim-berlitic samples collected in 2001 have been determinedby the Geological Survey of Finland, and these data willbe used in modelling the geophysical response obtainedin the airborne surveys. It is planned to collect furthersamples for petrophysical measurements in 2002.

Conclusions and plans for further workField work in 2000 and 2001 in poorly known areas hasadded approximately 50 samples of kimberlitic dykesand more than 300 kimberlitic boulders to the existingdatabank.

Niobium in the fine fraction (< 0.1 mm) of streamsediments has proved to be a convincing ‘pathfinder’element for kimberlites and lamproites in the Sisimiut–Sarfartoq–Kangerlussuaq region. Nb anomalies and thepresence of high-pressure chromite in the stream sedi-ment samples in the eastern and south-eastern part ofthe study area warrant follow-up.

The different xenolith types from the locality A dykeand their distinctive mineral compositions clearly demon-strate the heterogeneous character of the West Greenlandlithospheric mantle, with a vertical zonation of depletedand metasomatised zones beneath the Archaean craton.Temperature and pressure calculations suggest that allxenoliths were derived from within the diamond sta-bility field at a depth interval of 49–69 kbar, corre-sponding to approximately 150–215 km.

Although magnetic data have been used with somesuccess in the search for kimberlitic dykes in theKangerlussuaq region, there is a need for establishinga database with petrophysical properties of both kim-berlitic dykes and host rocks in order to fully utilise thegeophysical data available. Magnetic profiling at ground

64

0.01

0.10

1.00

10.00

100.00

La Ce Nd Sm Eu Gd Dy Ho Yb Lu

REE

Con

cent

ratio

n/C

1 ch

ondr

ite

464002.2-GT464006-GT464003-GTLh-1-GTLh-2-GTHz-GT

Fig. 8. Chondrite-normalised REE patterns of garnets in xenolithsfrom the locality A dyke (this study) and of garnets in Sarfartoqxenoliths (Garrit 2000). Laser ablation ICP-MS analyses, samesources as in Fig. 7.

55476554505544055433

55428 55424

55420

55416 55412 55408 55404 55399 55392

1 km

nT

Fig. 9. Magnetic total field anomaly oflocality J (Fig. 2), where a 5 km long and20 m wide kimberlitic dyke shows up aspronounced linear feature. Data fromhelicopter-borne DIGHEM survey carriedout for Dia Met Minerals Ltd., reproducedwith permission.

GSB191-Indhold 13/12/02 11:30 Side 64

level should be conducted at selected kimberlitic dykeoccurrences in future field work.

Exploration companies have produced a large vol-ume of data relevant to diamond exploration since1995. The data include mineral analyses of heavy min-erals recovered from till samples, dyke and boulderdistribution maps, helicopter-borne geophysical sur-veys, results of diamond testing of mini-bulk sampleddykes, drill logs, etc. Compilation and publication ofall non-confidential company data submitted in assess-ment reports to the Bureau of Minerals and Petroleum,Government of Greenland, is planned for 2002. A com-pilation of this type covering the North Slave Craton inNunavut, Canada (Armstrong 2002) has received verypositive response from industry.

The ultramafic alkaline rocks observed within theregion north-west of the Sarfartoq cone sheet structure(Larsen 1980) have not been proved to be diamond-iferous. However, the finds of diamonds in kimberliticdykes of the Torngat orogen (Derbuch 2001) – theCanadian counterpart of the Nagssugtoqidian orogen –is an encouragement for further exploration in WestGreenland.

AcknowledgementsMajken Poulsen, a trainee funded by the Bureau of Minerals andPetroleum, Government of Greenland, is thanked for cheerful com-pany and valuable assistance during three weeks of the 2001field work. Berit Wenzell is thanked for assistance during elec-tron microprobe analysis, and Jørn Rønsbo for access to themicroprobe laboratory at the University of Copenhagen. TheDanish Lithosphere Centre, funded by the Danish NationalResearch Foundation, supplied microscope and computer facil-ities for H.H., and the xenolith study is funded by a CarlsbergFoundation grant to H.H., and by GEUS. Dia Met Minerals Ltd.is thanked for permission to reproduce the magnetic anomalymap of Fig. 9.

ReferencesArmstrong, J.P. 2002: Diamond exploration data, North Slave Craton,

Nunavut. NU Open Report 2002-01, 1 CD-ROM. Iqaluit, Canada:Department of Indian Affairs and Northern Development.

Derbuch, H. 2001: Twin Mining’s Torngat results advance project.News Release, 3 pp. Toronto, Canada: Twin Mining Corporation(issued 5 April, 2001).

Ferguson, J. 2001: Citation confirms large kimberlitic dyke inGreenland. News Release 2001-8, 1 pp. Vancouver, Canada:Citation Resources Inc. (issued 27 September, 2001).

Garrit, D. 2000: The nature of Archaean and Proterozoic litho-spheric mantle and lower crust in West Greenland illustrated

by the geochemistry and petrography of xenoliths from kim-berlites, 289 pp. Unpublished Ph.D. thesis, University of Copen-hagen, Denmark.

Griffin, W.L. & Ryan, C.G. 1995: Trace elements in indicator min-erals: area selection and target evaluation in diamond explo-ration. In: Griffin, W.L. (ed.): Diamond exploration into the21st century. Journal of Geochemical Exploration 53(1–3),311–337.

Griffin, W.L., Gurney, J.J. & Ryan, C.G. 1992: Variations in trap-ping temperatures and trace elements in peridotite-suite inclu-sions from African diamonds: evidence for two inclusion suites,and implications for lithosphere stratigraphy. Contributions toMineralogy and Petrology 110, 1–15.

Griffin, W.L., Shee, S.R., Ryan, C.G., Win, T.T. & Wyatt, B.A. 1999:Harzburgite to lherzolite and back again: metasomatic processesin ultramafic xenoliths from the Wesselton kimberlite, Kimberley,South Africa. Contributions to Mineralogy and Petrology 134,232–250.

Gurney, J.J. 1984: A correlation between garnets and diamondsin kimberlites. In: Glover, J.E. & Harris, P.G. (eds): Kimberliteoccurrence and origin: a basis for conceptual models in explo-ration. Publication 8, 143–166. Perth: Geology Department &University Extension, The University of Western Australia.

Gurney, J.J. & Zweistra, P. 1995: The interpretation of the majorelement compositions of mantle minerals in diamond explo-ration. In: Griffin, W.L. (ed.): Diamond exploration into the21st century. Journal of Geochemical Exploration 53(1–3),293–309.

Haggerty, S.E. 1995: Upper mantle mineralogy. Journal of Geo-dynamics 20, 331–364.

Hoal, K.E.O., Hoal, B.G., Erlank, A.J. & Shimizu, N. 1984:Metasomatism of the mantle lithosphere recorded by rareearth elements in garnets. Earth and Planetary Science Letters126, 303–313.

Keating, P. 1995: A simple technique to identify magnetic anom-alies due to kimberlite pipes. Exploration and Mining Geology4(2), 121–125.

Larsen, L.M. 1980: Lamprophyric and kimberlitic dykes associatedwith the Sarfartôq carbonatite complex, southern West Green-land. Rapport Grønlands Geologiske Undersøgelse 100, 65–69.

Larsen, L.M. 1991: Occurrences of kimberlite, lamproite and ultra-mafic lamprophyre in Greenland. Open File Series GrønlandsGeologiske Undersøgelse 91/2, 36 pp.

Larsen, L.M. & Rex, D.C. 1992: A review of the 2500 Ma span ofalkaline-ultramafic, potassic and carbonatitic magmatism inWest Greenland. Lithos 28, 367–402.

LeBas, M.J. & Streckeisen, A.L. 1991: The IUGS systematics ofigneous rocks. Journal of the Geological Society of London148, 825–833.

Macnae, J. 1995: Applications of geophysics for the detectionand exploration of kimberlites and lamproites. In: Griffin, W.L.(ed.): Diamond exploration into the 21st century. Journal ofGeochemical Exploration 53(1–3), 213–243.

Mengel, F. et al. 1998: Archaean and Palaeoproterozoic orogenicprocesses: Danish Lithosphere Centre studies of the Nagssug-toqidian orogen, West Greenland. Geology of GreenlandSurvey Bulletin 180, 100–110.

65

GSB191-Indhold 13/12/02 11:30 Side 65

Menzies, M.A. & Hawkesworth, C.J. (eds) 1987: Mantle metaso-matism, 472 pp. London: Academic Press.

Mitchell, R.H., Scott Smith, B.H. & Larsen, L.M. 1999: Mineralogyof ultramafic dikes from the Sarfartoq, Sisimiut and Maniitsoqareas, West Greenland. In: Gurney, J.J. et al. (eds): Proceedingsof the VIIth International Kimberlite Conference 2, 574–583.Cape Town: Red Roof Design cc.

Nimis, P. & Taylor, W.R. 2000: Single clinopyroxene thermo-barometry for garnet peridotites. Part I. Calibration and testingof a Cr-in-Cpx barometer and an enstatite-in-Cpx thermometer.Contributions to Mineralogy and Petrology 139, 541–554.

Nixon, P.H. (ed.) 1987: Mantle xenoliths, 864 pp. Chichester:John Wiley & Sons.

Nixon, P.H. 1995: A review of mantle xenoliths and their role indiamond exploration. Journal of Geodynamics 20, 302–329.

Olsen, H.K., Jensen, S.M., Schønwandt, H.K. & Helmstaedt, H.H.1999: Review of diamond exploration in Greenland. NorthAtlantic Minerals Symposium, Dublin, Ireland, 19–22 September,1999. Extended abstracts volume, 166–168.

Ryan, C.G., Griffin, W.L. & Pearson, N.J. 1996: Garnet geotherms:pressure–temperature data from Cr-pyrope garnet xenocrystsin volcanic rocks. Journal of Geophysical Research 101,5611–5625.

Scott, B.H. 1981: Kimberlite and lamproite dykes from Holsteins-borg, West Greenland. Meddelelser om Grønland Geoscience4, 24 pp.

Secher, K. & Larsen, L.M. 1980: Geology and mineralogy of theSarfartôq carbonatite complex, southern West Greenland.Lithos 13, 199–212.

Shimizu, N. & Sobolev, N.V. 1995: Young peridotitic diamondsfrom the Mir kimberlite pipe. Nature 375, 394–397.

Steenfelt, A. 1991: Reconnaissance geochemical mapping of south-ern West Greenland from 62°30´ N to 64°00´ N – 1991 results.Open File Series Grønlands Geologiske Undersøgelse 91/9,21 pp.

Steenfelt, A. 1997: High Ba–Sr–P–REE components in EarlyProterozoic magmas within the Nagssugtoqidian orogen, WestGreenland. Posters and Abstracts. 9th EUG symposium. TerraNova 9, 360 only.

Steenfelt, A. 1998: The Nagssugtoqidian orogen imaged by geo-chemical maps. In: Wilson, J.R. (ed.): 23. Nordiske geologiskevintermøde, Aarhus, Denmark, 13–16 January, 1998. Abstractvolume, 234 only. Aarhus, Denmark: University of Aarhus.

Steenfelt, A. 1999: Compilation of data sets for a geochemical atlasof West and South Greenland based on stream sedimentsurveys 1977 to 1997. Danmarks og Grønlands GeologiskeUndersøgelse Rapport 1999/41, 33 pp.

Steenfelt, A. 2001a: Calibration of stream sediment data fromWest and South Greenland. A supplement to GEUS Report1999/41. Danmarks og Grønlands Geologiske UndersøgelseRapport 2001/47, 43 pp.

Steenfelt, A. 2001b: Geochemical atlas of Greenland – West andSouth Greenland. Danmarks og Grønlands Geologiske Under-søgelse Rapport 2001/46, 39 pp. + 1 CD-ROM.

Steenfelt, A., Dam, E. & Nielsen, J.P. 1993: Reconnaissance geo-chemical mapping of map sheets 67 V.1 and 68 V.1 (66° to68° N, 51°25´ to 54°W), West Greenland. Open File SeriesGrønlands Geologiske Undersøgelse 93/1, 18 pp.

van Gool, J.A.M. et al. 2002: Precambrian geology of the northernNagssugtoqidian orogen, West Greenland: mapping in theKangaatsiaq area. Geology of Greenland Survey Bulletin 191,13–23 (this volume).

66

Authors’ address

Geological Survey of Denmark and Greenland, Øster Voldgade 10, DK-1350 Copenhagen K, Denmark. E-mail: smj@geus.dk

GSB191-Indhold 13/12/02 11:30 Side 66